CHEMISTRY: W. F. LIBBY

consider the relative motion of three hydrogen atoms. They formed the "skewed"co-ordinate system in which the potential energy surface was plotted. Our setof relative co-ordinates generalizes the "skewed co-ordinates" to systems of Nparticles. These co-ordinates should be useful in molecular collision problemsand molecular rotation-vibration problems. Their principal use, however, willprobably lie in the prediction of the variation of chemical reaction rates withisotopic masses, since the zero-point energy in the activated complex and also thetransmission coefficient depend rather sensitively on the masses which appear inthe relative co-ordinates.

The authors wish to thank Elizabeth S. Hirschfelder for her valuable help duringthe course of this work. Also, John Dahler wishes to thank the National ScienceFoundation for financial assistance.

* This work was carried out at the University of Wisconsin Naval Research Laboratory underContract N7-onr-28511, with the Office of Naval Research.

t Present address: Institute of Theoretical Physics, University of Amsterdam, Amsterdam-C,Netherlands.

l H. Eyring and M. Polanyi, Z. Physik. Chem., B, 12, 279, 1931; J. 0. Hirschfelder, dissertationPrinceton University, 1935; S. Glasstone, K. Laidler, and H. Eyring, Theory of Rate Process(New York: McGraw-Hill Book Co., Inc., 1941), p. 100.

RADIOACTIVE STRONTIUM FALLOUTBY DR. W. F. LIBBY

COMMISSIONER, UNITED STATES ATOMIc ENERGY COMMISSION

Communicated May 2, 1956

CONVERSION FACTORS

"Average soil" = 20 gm Ca/ft2 in top 2.5 inches1 MPC unit in any medium = 1 Mc Sr9"/kg Ca

= 2,200 dpm Sr9"/gm Ca1 megaton fission distributeduniformly over entire earth = 0.0009 MPC unit

= 0.5 me Sr'0/mi2

I. EXPERIMENTAL MEASUREMENTS

Strontium 90 is of particular importance among the fission products becauseof chemical and physical characteristics which result in comparatively high retentionin the skeleton. These are chemical similarity to Ca, an element essential to bothplants and animals; an average life of about 40 years; and a low rate of eliminationfrom the skeleton. On the basis of studies of the comparative effects of Sr90 andRa226 in experimental animals and of the effects of Ra in humans, the generallyaccepted maximum permissible body burden (skeletal content) of Sr90 in adulthumans is 1 microcurie. Since the body of the average adult contains about 1,000gm. of Ca, this is equivalent to saying that the maximum permissible averageconcentration of Sr90 in the adult skeleton is 1 lucI1,000 gm. of Ca. For purposes

VOL. 42, 1956 365

Dow

nloa

ded

by g

uest

on

July

1, 2

021

CHEMISTRY: W. F. LIBBY

of this discussion, this ratio of Sr90 to Ca, in whatever medium it may occur, isdesignated an "MPC unit." One MPC unit of Sr9O in the human body isconsidered to be safe-a significant risk occurring only at much higher dosages.'The majority of analyses for Sr90 encountered in this work were of the order of afew thousandths of 1 MPC unit. For purposes of orientation, it is helpful toremember that 0.001 MPC unit corresponds to 1/1,000 lc of Sr90/kg of Ca, or 2.2dpm of Sr90/gm of Ca. The small weights of Sr90 involved in both the radio-strontium and normal strontium being considered, and the similarities of theelement to Ca, justify the assumption that its distribution in the body will followthat of Ca in a general way.Two megatons of fission will produce 1 millicurie (mc.) of Sr90/mi2, if the fission

products are uniformly distributed over the earth's surface. If this amount of radio-activity is mixed with the available Ca in the soil, an average of about 20 gm/ft2 inthe top 2.5 inches of soil, the specific radioactivity produced is 0.0018 MPC unit.It is observed that most of the Sr90 fallout is concentrated in the top 1 or 2 inchesof soil. For example, in Tables 1 and 2, which show the Sr90 burden in the fall of1953 in the soil of twelve farms in the Wisconsin-Illinois area as well as in thealfalfa and the milk of the cows fed thereon, we note that the top inch of soil containsabout 56 per cent and the next 5 inches contain the remaining 44 per cent of thetotal Sr90. Recently some evidence has been discovered that the radiostrontiumfinds its way to greater depths.2 In Table 2 data are given for Iowa soil collectedin 1937 which, as expected, shows no Sr9". The average available Ca content of thedomestic soils was 8 + 1 gm Ca/ft2/in., the average fraction of total Ca ex-changeable was 68 i 3 per cent, and the average Sr90 content was 4.7 ± 0.4 mc/mi2.

TABLE 1BIosPHEoRE Sr'0 ASSAYS* (WISCONSIN MILKSHED-PRE-CASTLE, OCTOBER, 1953)

(Values Are Given in Terms of 0.001 MPC Units, Except where Noted)TOTAL Sr90

SoilFARMt 0#-i' 1'-6' (MC/Mi2) Alfalfa Milk

Grabow, Wisconsin 26.2 6.7 4.5 12.8 1.7Oliver Swain, Wisconsin 7.4 2.2 3.1 5.3 1.3Swanson, Illinois 15.8 2.5 9.2 7.1 1.2Holcomb, Wisconsin 8.7 1.8 5.1 8.3 1.6Lewke, Wisconsin 10.2 2.9 3.5 20.9 2.3Premo, Wisconsin 13.1 2.5 3.8 4.1 0.7Kurpeski, Illinois 16.3 6.6 4.0 7.4 1.3Austin, Illinois 22.4 4.7 4.7 5.0 1.8McKee, Illinois 8.1 0.9 6.3 14.8 1.4Blomberg Illinois 1.7

ho - U) hiV 00 0 NOmC., ( C CoE-4 v c eq eq 00 eq N co co CYD eq m eq N 0

'-'eq Qj - - CY eq - "4 " eq q - _- eq cq eq

eq

eq

-.

0)

a ,s 00

0,

~~~~~c

CD

cS :S

vOO X

., A

xa, +

o^ ,$g °

to N

00

-eq

00

40

eq

00

4

1-'

V4

001'4

0

0

ho.4

"di eq 00 uhoo ei ko o. .

4

0000

t. 00 eq

00 "4 000 _ Ula U3 Co Co 1'eq Co eq cq ch ho eq

- _ - eq

t'.hi hi t -4 to Co C C00 eq eq hoeq eq 0 "4 -" 0 e-eq C 00

co0000

N 00 0 C C co ho N - cho to 0 Co C 004 h0 4 N 00 N eq 4 t' CD 0 4 *"4 co co co cq cq -4

ho o 4 CoD 00 eq - ho ho 0) 0 - co co ho 0)CVo 00 O 00 00 0 " hi N b N ho1 + ho

o hi kO 0 hi, ho 0 0 ho ho ho 0"4 0 0 N~ Nz Nl 0 00 0) 00 00 0"4 "-4 "4

ho h o 00 0 to O Co N 00 eq cq t(0 0 0) Co Co ~4 ~ 4 Nz o' 6 0 h

"-4 "4 " "4-1 "4

hO h0 0 ho h10 00 ho ho Co

O 0 0 ho00 0 00 0

h

b.ho hiCC6 4 C6 k

to hi ho a 0 0 0° o00 b 0 s s q O 06 C;

0 00 + ho 0 + ho N 0 0 o eq a o eq 0CO o N 00 0 0 h. N o q h.o o CoP4 o -ooko4o oetc

O hoc 00Co 4 co eq eq Co Co eq eq00 14 0 "4 0 "4 0 "4

00 h 0 "14eq eq Col Ci Co Co"-4 "4 0 ,.4 0 "4 0

o3 o ho hoi ho

035 A 10c 00p sa q eq Cq eqw E-4 N,

00 0

C C CO c o C Co co Co coho ho ho ho ho to to ho o00 00 00 0 00 0 0 0 0 0eq eq eq eq eq CIO Co Co co Co00 00 00 0 00 00 00 00 00 00

N. 00 00 0 1- eq Co) uho to hoho co ho N N- -co CD Ch ho to ho co ho ho co

4 -4 4 44 "4 "4CIO C

ho to o ho ho a ho ko

46 *_ *_ 4

tO 4 o

tN 00

hou ho

.e .d .<

* - . -

00 00

Z4,9 0 ClZ 0Z

4ea *ad *0d*o2 $ S.; .r43

367

0o 0 "co CO -1N, N, 000 00 "4

ho-

-11-

ho hoh0ao

0) 0)a, 0,

0 NO o

44"

l00

CD

do e

coto

ho

.4

to

01)0,

0

6z0)

00

"0Agr as

Dow

nloa

ded

by g

uest

on

July

1, 2

021

04~~~~~~1

1Pq14 9 0) l

.-P4

0 04 0

)H 0

togX

o..*

0. 0)

04

tied d 0

0 . 0.0 >b b °

'0-]t X 4

4)~

co~

e m

0.o4): S

X0-0t~m O s sa~~C

H C

4)-4

@7

_p¢

0

1. m .

o CO

0

00

00

0O

eq

0000 CO 0)co CY) 0

0 0 00 eqN 0q .-4 ei eq . 00

mCO

-0

Cm O0e

1. -

0) 0)

e CO COeq _ _

0 0 0

10 1Dtl .i .4

eq eq

0 0 0.1 .1

0

00 0) to

q

10f- t-- t--o eq

eq eq

C

0 CO COeq

0:: 1

0) 0 -4

10 CO

Ce

0 0CO el

O

000)

'

t- C OD00

eq eq4 eq4eq

0 N-4

0

000

00 4 0 00eq eq oo

eq eq

0 0 0 0

0

0 0

01

0o _ eq eq eqCO eq cq _ co

_4 P_*

14 II4 C6 4

_4

CO

1-

2

0 0

~0 0 T

14 ko U3eql 00

CO

_ _. "I

10 10

co co coco CV2 V:

to ul ko

CDi0

Co co 00 c0eq eq eq co CO V e"4 "4 O _

co co

C1

0

CO CO CO1_

eq eq eq

0 0 0

CO CO

_. _041.

+4 U1 t- 00 0 0oo 00 00 00 00 00C0 00 0oCO CO CO CO CO

16

M4) 4; g

Z>4 >4

b OD tD OD V G

0D 4m04 be

3

3.68

368

4)

-H

4)

)

w D

V%

a0

ec

0 0

C , +2

0. 0.

AH n o

4)s ._ _0o 00O0GS S

d b ID C

4)@ "4CO.

2

c~o

0

CD

)4)0

O C)

O 0

4).m

*03

-"

tzo

asas00:

o o

-01 4

..

0 0 tt4p *.a_1010

s4)4

00o

11 ,C" 4)4).

4-u:

4)4 .o

* 00

00>chm

4 eq

~ 0.

0o 0 .ow

0. ce.0

¢.

v

Dow

nloa

ded

by g

uest

on

July

1, 2

021

VOL. 42, 1956 CHEMISTRY: W. F. LIBBY 369

The amount of radiostrontium found in humans is shown in Figures 2a and 2b.The data show that the present Sr90 content probably averages somewhat lessthan 0.001 MPC unit in young people. Apparently a number of barriers protectthe human skeleton from this fallout radioactivity.

01

I.

0.007_

0.006 _

0.005 _-

0.004 _-

0.003 _-

0.002

0.0010 -

0.000 L

0

: *S.

0

0

I 019S3 1954

DATE

FIG. la.-Wisconsin cheese-Sr9° content

19SS

* C00CAGO MEASUREMENTS 8612o LAONT MEASUREMENTS 2- X

0.004w a

I.3 a4 A0.03 _ *

a**

0.006'953

iZ

.Z04XI"

*e a* OITALY

I5IaI

1954

FIG. lb.-Foreign cheese-Sr' content

Measurements have been made on animals, principally cattle and sheep. Thesedata are given in Figures 3a and 3b. We see here; that the contents are muchhigher than those for milk and human samples, apparently due to selective dep-osition of Sr in the animal bones, which protects the milk and thus human bone.

'AS.0

zI-

vILx

'955

Dow

nloa

ded

by g

uest

on

July

1, 2

021

CHEMISTRY: W. F. LIBBY PRoc. N. A. S.

Data for foreign soil samples collected just before the Castle test series are pre-sented in Table 3 and Figures 4a and 4b. From these data we deduce that theband around the earth bounded by the latitudes 600 N. and 100 S. shows a deposi-

* CHICAGO LABORATORIES *-12HEALTH AND SAFETY LABORATORIES NYO 26.29

: (NEW YORK CITY SMPLES ONLY)

0~~~~~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~

U ILo

0.006 -A

T S

_00.005_ < *

004 V 0 0

~~~~~U0

o o00 0~~~~~~~~~~~*UTAH* JANES^VLLC 000°o° 0 0WISCONSIN 08M 0

oo *0 NEW YORK CITY

1953 19t4 195

DATE

FIG. lc.-U.S. mllkR-Sr9 content

0 CHAGO LABORATORIES 12

0

X J~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~-0.005 '

0.000 _

8.003 *<

_.002 Th11 *-

1953~~~~z 5 195

~ ~ ~ IG Id.F 1eilmilr otna ~~~FRANCEITALY

0.001 zzM0TURKEY - R ~~~~~~~~~~~PERTH,AUSTRALIA

1953 1954 1955

FIG. id.-Foreign milka-Sr' content

tion of 0.8 megatons (MT) equivalent of SrO0, in addition to some 0.4 MT whichappears to be nearly uniformly deposited, as would have been expected from a slowdeposition frojn a large stratospheric reservoir. No evidence for longitudinal varia-tion is apparint in Figure 4&.

370

Dow

nloa

ded

by g

uest

on

July

1, 2

021

~. -H -H'ho -W

00I

~~~~~~~~ ~~~*eq

C;,q eqcqoCO

goN

o tCO~~~~~~~~~k

r) 0 -

z t0-

00

Eqzz

CDCz 1*g.N . L

X) GX O co m

zX coa

00 0

1 > - >

05

0 ?0LO Lo

X ~ ~~~ ~ ~~~~¢ m N

oooo_ oooo-Vl VIVI

10C~000000000000co. .. .. .0 000C oooao oo-oCoooo o o orio o o oVl-H H -H H -H H -H H -H H -H H 4i H 41 + -H 4 -H *. -H -f VIVI -H

esq to t -_ce1-00_CD 0 C 0Cmt O 10" -cN N m cm C:D0 N ulz a k oo0 OM kza o* * *COCO * -4

00-00000---0 CO -00 eq

00 N c0t.C ' 0 b. b. CO C

*0O

Vl010 C oo-(.0 0 C 0t.. e CO

-H + VI -H + -H -H1 v00

C . . .-0_ 00o -e

O Co co eq O0) t.-0COc o eq 0 ' CLO CO0 ) bt0-0

C 0)_ -4 0 - -4 _- 0 - t- 0 N t- 0 0 N _- co bo ' _

CO o~ 1_.u t 0cioo _0o C6 C ec ct M o

w~~~~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ - v_4 co_0 ml _

eqeqCOCO-COlacCOC eqCiqCO0O! Ce000u: eC 0 eq 0C qc0o 10CO eCOCoo0

UO V -4 co -4 _OVCeOe c co

. .

0 t- t--Ob m 00cq oLee co - 0_ 9o tso c ~06 1 Co' 00 06 -4 Co'

oD 0 mce N WC ko N NO O ooo 0 cV COko koNo 00o O es

OIA Oe Co eq - 00 O -_ 0 0co co I-COco cD o O N ko LO co C*ce _ I_ 4_ 4 4 04_ 4C _ 4 _4 _4 _4I"!0)-C9O C! qCOC00! qeq~C!C! iO0CO ll CO1 0-I!1il ~ 10 0 t

o 0 t co m 0m m d m m. o0 00 0 VCO - C0-0C VcOctoeto 0Oe 0-0 CO ceq -00C V00 m)4 o-C _oo M 0CO _

00 00 IV v v -0 .4 14 00 00 0 00 t- oO 0 00 t- 0 00 uo I 00 00 0o (M 0

o meqo CO 00 -k to aco Ieq % o0 e co0 c Mc_CoCOO' o M t eC-Z 0 e° .0 0 oOco, r. C6

00COCOCO~COCO0)t-.CO"4'0 V14 "4 O C O 0

_-_ _ cc _ _ 00to t- t w000eq eq LOeq - Cq eq eq t-to-.-t- t- $D

> +e e e tv .4 v kotoae ko u" V V U' Lot4X0 10 10 - - C--- 101010

V eq eq e a N q 4 NCCCq0-00

m m ell N co m m c c cOe qeqeeq eN V V _

0 0 00 o o000 o 0 00 010 0 01000 00 0000 000 000 0t-.a-.0c0b b e O%e-q *o0o t _ t

*z z z **z z z z z z z z z z Z z z z z z0zo z z o o o 0 0o 0 0 0 0 0 0 0 0 o o o o

O O ooQeq o o o o Cq C 0 0eq to Lo 10qocq Ne" eqeeo to10ccqeqc m k k u o 010 c co CO UO 10 ko U to vl

371

10iW-

CO 0 C104 CO -oo"

a.

eO 0* a-c...o 0O cO Nzt_0tQ O

0.

00

CO

0O 00 00 00 0000

tc

kOOaOI"t¢-'-

0g- - 0

00O 0 q .Cq CS < > > 00m

s ad

0

.O.c

t Z

SOC0O ,l~ 4)

D 3- Q -@QC CO

la 0 0V ¢

, cocoaaB

6ZZ

COC°O- qe eqs eqsa

O O O OOO *

000 LO 0 00

4o 4 co co u

eqeqX O _

Dow

nloa

ded

by g

uest

on

July

1, 2

021

CHEMISTRY: W. F. LIBBY PRoc. N. A. S.

(MEASUREMENTS BY CHICAGO LABORATORIES 8-12)(ERROR LESS THAN 0.00005, UNLESS OTHERWISE INDICATED)(PARENTHETICAL NUMBERS ARE NUMBER OF CASES USED FORTHE AVERAGE)

CHICAGO

.. I. i@

-is.X'$.

z

8v

li

u-4 v~~~

CHICAGO* 0f mgmA)AC

LI two" byf

a

+;

19S3 1954

FIG. 2a.-Human stillborns

WTE

FIG. 2b.-Human bone-Sr98 content

372

.002

0.001 _

I-

nI

Iff

Dow

nloa

ded

by g

uest

on

July

1, 2

021

*CHICAGO DATA 8-12o LAMONT DATA 2, 25

FIG. 3a.-Calf bones

0.0 l.0zZ UU

,o "I1 OZXz -c 00D

0 z0 > eo0^^ 04oo,^^

FIG. 3b.-Sheep bones

0 .

*0

0 0 * 04?1954

DATE

I.-0k..1Z95zW3

.coc 0

I 1Gi..S3

0.07

O."oI

0.050

1 0.040

I-z

u

CL

0.030

0.020

0.010

0.000

0.070

0.060

0.050

. 0.040

I2

0 0.030

0.020

0.10I

@0

* 0V

1953

.

Dow

nloa

ded

by g

uest

on

July

1, 2

021

CHEMISTRY: W. F. LIBBY PROC. N. A. S.

The actual heights of rise of the bomb clouds are the basis for the assumptionthat all distant fallout from megaton weapons occurs from a stratospheric reservoir,while that from those of lower yield occurs from the troposphere. Actually, theheight of the tropopause varies with the season, so that the season must be con-

I Q(2T = 19.o5/.i2 OVER WHOLE EARTH)

Xi0

- 0IMEAONTOA

>

I' Ia

0.4 MEGATON TOTAL.

'ai

90 to 70 GO 50 40 30 20 10 0 10 20 30 40 sR 60 70 to ".NORTH LATITUDE SOUTH

FIG. 4a.-Latitudinal distribution of foreign pre-Castle Sr"0 fallout (soil assays)

*

0 M

I0 II

zI..0 (00 i0 @ * 0 .1~~~~~~~~~~~0 !

AVERAGE (AREA WEIGHTED) a0- - ...--------G-- -0---0w- I- -o

0 0* 00oa

0 eANTARCTIC SNOWS (el TABLE an< 8

* @~~~~0

0I

i _ _ 8___ .aI.-990 lOR oo N 40 20 0 20 40 0 o INNl, 120 140 I" I1

"EST LONGITUDE EAST

FIG. 4b.-Longitudinal distribution of pre-Castle Sr90 soil data from foreign samples.

sidered in the assignment. During the Pacific tests it has been near 55,000 ft.;hence our classification has validity in this respect.Two of the most important data in Figures 4a and 4b are those from the ant-

arctic series. The samples were snow cores, collected for the Chicago Sunshine

374

2

.i

i

a.AA14

A

CHEMISTRY. W. F. LIBBY

Laboratory and for cosmic-ray tritium analysis,3 by Mr. Paul Humphrey of theUnited States Weather Bureau in January and February, 1955, at Admiral ByrdBay (69°34' S.; 00°41' W.), at Atka Bay (70°35' S.; 08006' W.), and at LittleAmerica III. The data are given in Tables 4 and 5. Table 4, Part A, gives theSr90 and T contents of surface snow at four locations. From the T concentrationsand the expected T production rate in this region" (T produced in the Castletest itself was precipitated out in a few weeks and never entered the SouthernHemisphere appreciably, because of the large amount of water taken into the cloudwith which it became mixed6), we determined the annual precipitation to be 7.8 2inches of water. This calculation, together with that for the January-February,

TABLE 4POST-CASTLE Sr0 FALLOUT IN ANTARCTICA

A. SURFACE SNOWBr98 T

Date Location (dpm/Liter) (Atoms/10"8 H's)January 15, 1955 Near Quonset, Little America III 3.2 0.3 14. 1 4 0.6January 17, 1955 One-half mile east, Little America III 3.1 :1 0.7 7.5 4: 0.6February, 1955 Atka Bay, 6 miles inland on shelf (70035'

S.; 08 06' W.) 5.3 :1: 0.5 19.2 :1- 0.8February 19, 1955 Admiral Byrd Bay (69034' S.; 00°41' W.) 2.0 A: 0.2 24 4i 5

Average 3.4 4 0.5 14 4 3B. Sr'* PRECIPITATION RATE IN JANUARY AND FEBRtTARY, 1955

Annual snow precipitation rate from T assay* and 0.59 T's/cm2/sec as the expected antarcticcosmic-ray T production.

p = 4-7 X 59 meters/yr = 7.8 : 2 inches of water.Sr's rate of precipitation:

3.4 4: 0.5 dpm/liter = 62 dpm/ft2/yr for 7.8 inches of *rater/yr = 0.8 4 0.2 mc/mi2/yr.* H. von Buttlar and W. F. Libb , "Natural Distribution of Cosmic Ray Produiced Tritium. II," J. Inorg.

and Nuclear Chem., 1, 75, 1955; L. i. Currie, W. F. Libby, and R. Wolfgang, PhI40. Rev., 101, 1557, 1956.

TABLE 5PRE- AND POS-CASTLE Sr9° FALLOUT AT ADMIRAL BYRD BAY (69°34' S.; 00041' W.)

(Collected 2/19/55)SNOW CORE

Age (Years)(7.8 Inches Depth Sr9* T Sr90 Rate*Water/Yr) (Ft.) Density (dpm/Liter) (Atoms/10O H's) (Mc/Mi2/Yr)

00.54 0-1 0.35 1.95 4 0.20 24 A 5 0.460.54-1.04 1-2 0.32 1.7 4 0.2 12.5 : 0.8 0.401.04-1.52 2-3 0.30 0.48 4 0.04 13.5 4 0.7 0.111.52-2.16 3-4 0.41 0.90 + 0.06 ... 0.21

* Assumed annual precipitation, 7.8 inches water/yr, on the basis of T contents of surface waters (of. Table 4).

1955, Sr90 precipitation rate of 0.8 mc/mi2/yr, are given in Part B of Table 4. Theresult for the annual precipitation agrees well with direct observation by the AtkaExpedition and by Mr. Humphrey personally.7 At Little America IV, directobservation showed that the floor of the tent projecting from the ice front wasbeneath only 7 or 8 feet of snow after roughly seven years. Since the density ofthe snow was about 0.35, this would correspond to about 5 inches of annual pre-cipitation. Other observations7 checked this general magnitude. In Table 5 acore taken at Admiral Byrd Bay was measured for both Sr'10 and T. From thesedata we observe the pre-Castle fallout rates of 0.11 and 0.21 Sr90 mc/mi2/yr. Thesurface rate at this site is 0.43, which is less than the general average in the area forJanuary and February, 1955, of 0.8 mc/mi2/yr-as shown in Table 4-hence it

VOL. 42, 1956 3X75

Dow

nloa

ded

by g

uest

on

July

1, 2

021

CHEMISTRY: W. F. LIBBY

may be that the pre-Castle values at this site are low also and should be increasedby the ratio 0.8/0.43, or by 90 per cent, to 0.2 and 0.4, respectively.The average Sr90 content of rain and snow in the Chicago area since the fall of

1952 was calculated by weighting each datum by the total rainfall observed in theparticular storm. The data on the Sr90 content of Chicago rain are given inFigure 5.8-12 It is clear that large fluctuations can occur in individual storms.However, these extremes were, in general, of low total rainfall; hence the effecton the average is small. It is interesting that the antarctic snows have about thesame Sr9O content as the average Chicago rain of Figure 5. We recall that pre-

AVERAGES WEIGHTED ACCORDING TO RAINFALL( 1953 POINTS HAD 5.1" TOTAL1954 POINTS HAD 19.8" TOTALAVERAGE ANNUAL IS 31.5"WHERE 4dpm/gaI. - lfc/mi2/yr.

<0 04

>~~~~~~~~~~~~~~~~

t~~~~~~~X

If4X I- X~~~I >

100 ~~ ~ Ju0

E*

Z so

~~~. z

* a4L hi;/}S. K 4 E *

S~~~

1 9S3 1954 1955

FIG. 5.-Sr90 content of Chicago rains

cipitation there is only one-fourth to one-fifth that in Chicago. According to themechanism espoused in this paper, these fluctuations are to be expected, becausethe fallout from the stratosphere is thought to be steady and continuing, and thewashing-out by rain is expected to carry down the fallout accumulated since thelast precipitation from the particular air mass involved. Many of the samplesgiven in Figure 5 and in Project Sunshine Bulletins Nos. 10 and 118-12 have beenmeasured for T as well;3-5 hence further correlations of the type described abovefor antarctic snow (Tables 4 and 5) can be made.

In Tables 1 and 2 the Sr90 contents of soils in the midwest region of the UnitedStates were shown to have an assay of 4.7 Mc/mi2 in October, 1953. The totalfrom rains in the preceding year was only 1.5 Mc/mi2, according to Figure 5; hencewe have to expect about 3 Mc/mi2 to have been deposited prior to Operation Ivyby direct fallout, most reasonably from tropospheric debris. The total fired in

376 P-Roc. N. A. S.

Dow

nloa

ded

by g

uest

on

July

1, 2

021

CHEMISTRY: W. F. LIBBY

Nevada prior to this time which would not have fallen out in the immediate vicinityof the test site was approximately 200 KT for the Operations Tumbler-Snapper andBuster-Jangle together. If this were all deposited in the United States, it wouldamount to about 7.0 mnc/mi2. The test series Upshot-Knothole, with approxi-mately 220-230 KT of distant fallout, and Teapot, with approximately 160-180KT, should have added correspondingly to the 4.7 mc/mi2 in October, 1953, andto the 3.0 mne/mi2 of stratospheric fallout for the intervening period, for a total ofperhaps 12 mC/mi2 expected in the spring of 1956 in the United States.The efficiency with which rain removes fallout from the air through which it

passes is probably high. One knows, on simple physical grounds, that as littleas 0.1 inch of rain will traverse at least 90 per cent of the air volume lying belowthe layer in which the rain originates, so that 90 per cent of all particles whichcan be swept up by a falling raindrop will be carried down by such quantities ofrain. On the point as to whether fallout is likely to be deposited by rain, we notethat G. H. Wilkening"3 showed that the decay products of the radioactive gasRn which in themselves are isotopes of nongaseous elements are found affixed toparticles of diameters between 20 and 800 angstrom units (0.002-0.080 ,)-asubmicroscopic range not at all unlikely for the radioactive fallout stored in thestratosphere.' The velocity of fall for such particles would be very small and inthis respect quite compatible with the long stratospheric storage times indicatedby the Project Sunshine data. Blifford, Lockhart, and Rosenstock'4 studied theconcentration of the Rn decay products in rainfall in the Washington, D.C., areaand concluded that rain was the mechanism by which the particles containing theseproducts were precipitated and that the average time the decay products spent inthe air before being precipitated was only 15 days, approximately. 0. Haxel andG. Schumann,"5 found this time in Heidelberg, Germany, to be about 4 days, andDamon and Kuroda'6 concluded that Blifford et al., were correct in attributing theprecipitation of the aerosol carrying the Rn decay products to rain, their conclusionbeing based in part on additional data that they had taken. The average timespent by water in the air was found by von Buttlar and Libby4 to be between 5and 14 days.For these reasons it seems very likely that rainfall or snowfall carries down a

major part of the fallout which comes from the stratosphere and probably is a veryimportant mechanism for that part of the tropospheric fallout material whichdoes not fall out in the first few hours or day or two after the detonations. Ofcourse, the whole question can be settled by direct experiment in which a correlationbetween rainfall and total fallout is sought. The present data seem to favor thehypothesis. This conclusion and prediction seem to be borne out by Table 6,which presents the total Sr90 content of the top 2 inches of typical United Statessoils collected in October, 1955, and leached at room temperature for 30 minuteswith 6 N HCl.

Table 7 shows data obtained in the Chicago laboratories on the Sr90 contents ofrivers and lakes. It is clear that these are much lower than those of the rain fromwhich they are derived. For example, from Figure 5 we should estimate that theaverage rain in the Chicago area in the summer of 1953 had a Sr90 content of about7 dpm/gal. From Table 1 for the domestic pre-Castle soil contents and fromFigures 4a and 4b for foreign pre-Castle soil contents, we estimate that the European

VOL. 42, 1956 377

Dow

nloa

ded

by g

uest

on

July

1, 2

021

HEMISTRY: W. F. LIBBY

rains averaged about 3 dpm/gal. The four rivers, Mississippi, Mosel, Seine, andDanube, show less than 5 per cent of this; hence we conclude that the Sr' in rainis removed by the soil before the water runs off to the rivers and lakes. This factagrees, of course, with the sharp localization of the Sr10 in the top 2 inches of soil(cf. Tables 1, 2, and 3).

TABLE 6*Sr9° FALLOUT ACCUMULATION IN ToP SOIL (0-2 INCHES) IN U.S. IN 1955

(Sampled September 23-October 20, 1955)Measuredin Soil

Station (dpm/Ft2)( 310t

La Guardia 350 ± 211 550 4 16

Binghamton 710 4 16Philadelphia 450 4 19Rochester 550 ± 20Jacksonville 470 4 20Atlanta 530 ± 12Detroit 640 ±t 21New Orleans 470 ± 14Memphis 900 A± 20Des Moines 540 ± 13Rapid City 1070 ± 21Seattle 400 ± 15Boise 1160 ± 23

Albuquerque 290 ± 20}Grand Junction 280 ± 14Salt Lake City 860 +: 18Los Angeles 120 i 12

Average 578or 7.3 mc/mi2

(Probable additional Sr"0 in lower layers and to be released by additional leachingprobably will raise this about twofold.)* Data by E. P. Hardy and R. S. Morse, of the Health and Safety Laboratory, New York

Operations Office, personal communication.t This datum was obtained by Dr. J. L. Kulp, Lamont Geological Observatory, Columbia

University. The procedure was different from that of the New York Operations Office(personal communication).

TABLE 7Sr99 CONTENT OF RIVER AND LAKE WATERS*

Sr90 Content Sr90 ContentLocation (dpm/Gal) Location (dpm/Gal)

Lake Michigan, October 27, Mosel River, Metz, September1953 0.39 i 0.08 7, 1953 0 ± 0.05

Mississippi River, Memphis, Seine River, Nogent, Sep-February 4, 1953 1.13 ± 0.16 tember 8, 1953 0± 0. 09

Mississippi River, St. Louis, Danube River, Ulm, SeptemberApril 17, 1953

CHEMISTRY: W. F. LIBBY

operation of the United States Weather Bureau, has been collecting fallout data"7by use of gummed papers of 1-ft.2 area which are laid flat for a certain time outin the open away from buildings. After the exposure, the paper is folded andmailed to the New York Operations Office in an ordinary envelope. Samples thuscan be collected cheaply, easily, and quickly from any populated area anywherethe postal service reaches. Most of the data so obtained have dealt with totalfallout rather than with Sr9O specifically, but many analyses for Sr90 have beenmade since Operation Castle. These are presented later.

(9?)

* 1953 DOMESTIC0 1953-195S4PRING. FOREIGN

: o-°t

zIu

_

- 0

o.olto-

0

00 0 0

o8o0 °0.-bPiol O

0

00 o

a 10 20 30 40SOIL CALCIUMCONTENT (/100g

o O I

SO 00 70

FIG. 6.-Top soil Sr9" concentration versus calcium content

The main question about the gummed-paper technique is its over-all efficiencyof collection. In order to determine this, the Health and Safety Laboratory hasconducted an extensive series of comparisons of the amounts of fallout by gummedpaper and a 12-gallon pot with an 18-inch vertical cylindrical wall placed imme-diately beside the paper. Some of the data thus obtained are given in Table 8.From them we deduce a collection efficiency of 69 ± 9 per cent (but we use 63per cent, since Mr. Eisenbud recommends this on the basis of more data and abetter statistical treatment).

VOL. 42, 1956 379

0.0+

.

Dow

nloa

ded

by g

uest

on

July

1, 2

021

CHEMISTRY: W. F. LIBBY PROC. N. A. S.

The data thus obtained- for the post-Castle Sr90 fallout rate in the United Statesand South America are given in Table 9. These are combined with those for otherareas, to give the world Sr9O fallout rates for September, October, November, and

TABLE 8GUMMED PAPER COLLECTION EFFICIENCY

(Relative to 12-Gallon Pot [18-Inch Vertical Wall,12-Inch Diameter; Cylindrical])

Gummed Paper Pot EfficiencyMonth (dpm/Ft2/Mo) (dpm/Ft2/Mo) (Per Cent) Reference

Mar., 1954 11.6 14 84 *Apr., 1954 15.2 31 49 *May, 1954 21.6 34 63 *June, 1954 10.7 9.2 116 *July, 1954 17.6 25 70 *Aug., 1954 13.5 7.7 176 *Sept., 1954 20.7 92 22 *Oct., 1954 3.1 11 29 *Nov., 1954 5.7 32 18 *Jan., 1955 9.0 9.9 92 tFeb., 1955 29.8 50.6 58 tMar., 1955 210 150 140Apr., 1955 44.9 79.5 57 §May, 1955 18.6 56.4 33 11June, 1955 12.4 51.7 24 11

Average 69 9#* Interim Sunshine Report, NYOO Report NYO4620, January 17, 1955.t Sunshine Report for January and February, NYOO Report NYO4643, April 21, 1955.Sunshine Report for March and April, NYOO Report NYO-4646, May 30, 1955.

§ Sunshine Report for May and June, NYOO Report NYO-4653, July 5, 1955."NYOO Report NYO-4623, January 18, 1955.# The figure of 63 per cent will be used for consistency. Mr. Merril Eisenbud (NYOO) rec-

ommends this on the basis of more data and a better statistical treatment.

TABLE 9POST-CASTLE FALLOUT IN U.S. FROM GUMMED PAPERS

(Taken at 63 Per Cent Efficiency [cf. Table 8])Sr90 Fallout Rate

Month Location (Mc/Mi2/Yr) ReferenceSept., 1954 Eastern U.S. (10 stations) 2.5 i 0.2 *Oct., 1954 Eastern U.S. (10 stations) 2.0 i: 0.4 *Nov., 1954 Eastern U.S. (10 stations) 2.0 i 0.4 *Dec., 1954 Eastern U.S. (10 stations) 1.4 i 0.2 tJan., 1955 Eastern U.S. (9 stations) 1.3 0.2 tSept., 1954 Western U.S. (20 stations) 0.9 i 0.2 tSept., 1954 U.S. (38 stations) 0.92 i 0.2 §Oct., 1954 U.S. (38 stations) 0.79 4- 0.2 §Nov., 1954 U.S. (38 stations) 0.95 i: 0.2 §Dec., 1954 U.S. (37 stations) 0.71 i: 0.2 11Sept., 1954 South America (12 stations) 2.1 4- 0.2Oct., 1954 South America (12 stations) 1.6 4- 0.2Nov., 1954 South America (12 stations) 2.4 i: 0.2

* Sunshine Report for March and April, NYOO Report NYO-4646, May 30, 1955.t Sunshine Report for January and February, NYOO Report NYO-4643, April 21, 1955.1 Sunshine Report for May and June, NYOO Report NYO-4653, July 5, 1955.Sunshine Report, August 1954, NYOO, HASL-S-2 (NYO-4656), November 12, 1954."Sunshine Report for July and August, NYOO Report NYO-4661, September 16, 1955.

December, 1954, presented in Table 10 and Figure 7. From these data we obtainedthese extremely important conclusions:

1. A Sr90 fallout probably derived from megaton weapons and nearly uniformover the world, except for local effects due to rainfall variations and to fallout from

380

Dow

nloa

ded

by g

uest

on

July

1, 2

021

VOL. 42, 1956 CHEMISTRY: W. F. LIBBY 381

submegaton weapons, seems clearly established. The fallout from the kilotonweapons lasts only a few weeks at most, since they involve only tropospheric

TABLE 10WORLD-WIDE Srm FALLOUT RATE FROM GUMMED PAPERS*(Taken at 63 Per Cent Efficiency [cf. Table 8]; Mc/Mi2/Yr)

Per CentTotal No.Earth's Stations

Area Area (Sept.) Sept., 1954ArcticNorth Tem-

peratePacificU.S.North TropicSouth Tropic

AreaArcticNorth Tem-

peratePacific (21

stations)U.S.North TropicSouth TropicSouth Tem-

perate (4stations)

6.5 5 1.2 *fi 0.4Oct., 1954 Nov., 1954

3.0 * 0.4 1.4 * 0.4Dec., 1954 Average0.9*0.4 1.6*0.2

10.9 14 1.7 * 0.24 0.68 * 0.24 1.4 * 0.24 0.6 * 0.248.0 2 0*40.6 0.6 * 0.6 1.9 * 0.6 1.2 * 0.61.5 39 1.3 * 1.4 0.9 * 0.14 1.4 * 0.14 0.9 * 0.14

18.5 8 0.7*- 0.3 1.5 * 0.3 0.7 *- 0.3 0.4 -* 0.325.5 9 3.2 *E 0.3 2.4 * 0.3 2.5 * 0.3 0.6 *E 0.3

Average (area weighted)Jan., Feb., Mar., Apr., May, June, July, Aug.,1955 1955 195 5 1955 1955 1955 19550.32 1.2 0.52 0.62 2.0 1.5 1.6 0.82

1.0

0.530.860.711.5

0.55 1.1

0.58 2.20.86 1.40.70 1.21.3 0.83

1.1 4* 0.120.9 *- 0.31.1 -- 0.070.8 *- 0.162.1 -* 0.151.5 -- 0.1

Average1.07

1.1 2.9 1.9 2.5 1.4 1.57

1.12.30.780.40

1.6 1.3 1.2 1.04.0 2.8 2.9 1.41.5 0.80 0.87 0.501.0 1.5 0.60 0.72

1.1 0.46 0.9 0.40 0.38 1.9

1.192.070.880.98

1.1 1.4 0.74Average (area weighted, except U.S. and North Temperate

omitted because of Teapot) 0. 95 * O. 1* Sunshine Report for July and August, NYOO Report NYO-4661, September 16, 1955, and personal com-

munication from Dr. E. C. Plesset, Rand Corporation.

-4

{+

TEMPERATE PACIe

I1f *.. rriM+,~. rolciRt 41

tOWJTRlOMC AXNTARCTICA

2

I-

3

.I

SI N a S N .. S S NH S S a M 0 S 0 N a S 0 0 N I r

FIG. 7.-World-wide Sr" fallout rates, September-December, 1954, (gummed paper at63 per cent efficiency except antarctic value, which was snow).

l

ARCTIC

4 0

4 p

Dow

nloa

ded

by g

uest

on

July

1, 2

021

CHEMISTRY: W. F. LIBBY

storage, but widespread fallout is found to occur at least 1.7 years after a megatontest series.

2. This average world-wide Sr90 fallout rate in the fall of 1954 and the springand summer of 1955 was 1.2 mc/mi2/yr (see Table 10).

Additional information available on the Sr9' distribution has been obtained byair filters operated at sea level, principally by the Naval Research Laboratory,measured in the Chicago Sunshine Laboratory, and by the Health and SafetyLaboratory.'8 The surface samples collected at Washington, D.C., by the NavalResearch Laboratory and measured at Chicago8-'2 are given in Table 11 and

TABLE 11AIR-FILTER DATA, WASHINGTON, D.C.

(Army Chemical Corps Type 5 Paper, 99 Per Cent Efficient Down to FewTenths Micron; 75 Per Cent Efficient Down to 0.01 Micron)

Equivalent* FalloutWashington, D.C. Sr'0 (dpm/106 Ft3) (Mc/Mi2/Yr)

April5-8,1953 18.6 ± 0.7 0.14October 2-6, 1953 41.1 ± 3.0 0.3October 6-9, 1953 30.5 ± 1.1 0.2October 12-15, 1953 70.4 i 12 0.5April 3-5, 1954 91.0 ± 7 0.7April 8-10, 1954 6.4 i 0.2 0.05April 10-12, 1954 258 ± 6 1.9April 12-14, 1954 65.5 4± 4.6 0.5April 15-17, 1954 11 ± 0.5 0.08April 17-19, 1954 21 ± 0.6 0.16April 29-May 1, 1954 32.2 + 2.6 0.2May 11-13, 1954 31.3 ± 2.2 0.2May 24-26 1954 216 ± 11 1.6June 1-3, 1954 68.3 ± 4 0.5July 16-17, 1954 73.5 ± 5.2 0.5July 26-29, 1954 48 ± 3.9 0.36November 1-3, 1954 120 ± 7 0.9December 1-2, 1954 103 ± 4 0.8January 3-4, 1955 281 ± 6 2.1February 5-6, 1955 127 ± 5 0.9February 10-12, 1955 241 ± 10 1.8February 22-23, 1955 202 ± 11 1.5March 3-4, 1955 270 ± 13 2.0March 7-8, 1955 394 ± 20 2.9March 13-14, 1955 267 ± 16 2.0March 16-17, 1955 310 ± 15 2.3March 22-23, 1955 393 ± 20 2.9

* 134 dpm/10 fts = 1 mc/mi'/yr (cf. text); (28300 ft3/ft2 = 106 ft3/35.5 ft2).

Figure 8. In order to correlate these data with the fallout rate, we recall that,as remarked previously, any rain of 0.1 inch or more probably will thoroughlywash down the fallout in the air below the layer at which the rain originates.Examination of the weather data for the Washington, D.C., area in the periodwhen the samples were taken shows that the average interval between rains was6 ± 3 days. Therefore, the Sr90 content of surface air should correspond to falloutfor this time on the average, and we would expect a fallout rate R (mc/mi2/yr) to

6correspond to a surface-air content of R X X 79 X 41 X 2.5 dpm/106 ft3,since 79 dpm/ft2 is equivalent to 1 Mc/mi2 and there are 41 ft2/106 ftI below thetropopause on the average at Washington, D.C., and 2.5 is the average ratio ofheight of tropopause to rain-bearing layers. The resulting rate of 0.70 ± 0.2

382 PROC. N. A. S.

Dow

nloa

ded

by g

uest

on

July

1, 2

021

CHEMISTRY: W. F. LIBBY

mc/mi2/yr is definitely low, compared to the rain result of 2.3 ± 0.2 mc/mi2/yrfor Chicago (Fig. 5) and of 1.5 ± 0.1 mc/mi2/yr for data involving gummed paper(Fig. 7 and Table 9). The uncertainty in the factor of 2.5 for the ratio of rain-bearing layer height is probably the principal uncertainty in the calculation of thefallout rate from air-filter data, though it may be that the perfect vertical mixingof the lower 40 per cent of the troposphere implicit in the calculation is an incorrectassumption, in that air right at the surface is cleaned to a considerable extent bysurface contact with vegetation and water and soil and therefore has less falloutthan the average for the lower troposphere.

(134dpn/l06*w.ft. - 5.,C/Mi2/y#)

AVE.: POSI CASTLEO .70 0.23PRETEAPOT-

0_

600~~~~~~~~~W~~~~0

TEAPOT

0 0 J F A MM J J A S ON 0 J F M A M J J A S 0 H D J

1954 DATE 1955

FIG. 8.-Sr' in air, Washington, D.C.

From the data given on rates of fallout, we calculate the average stratosphericresidence time, r = 10 5 years. The high-altitude data show a definite riseabove the tropopause. This is strong confirmation for the stratospheric storageand dissemination mechanism.

II. DISCUSSIONA. PREDICTED Sr90 FALLOUT

The stratosphere reservoir of Sr90 immediately after Operation Castle had beencompleted was about 12 Mc/mi2. The fallout rate of Sr9' corresponds to an averagestorage time of 10 4 5 years and essentially uniform world-wide dissemination.The radioactive half-life of Sr90 is 28 years, corresponding to an average life of 40years. Therefore, the Sr9O fallout rate from tests up to and through the Castleseries should be given by

R = ( 1 ) 12 1(1/10+1/40) mc/mi2/yr (1)

VOL. 42, 1956 383-

Dow

nloa

ded

by g

uest

on

July

1, 2

021

CHEMISTRY: W. F. LIBBY

where t is the time elapsed, in years, since May, 1954. This relation is given inFigure 9. From it, we can predict the tropospheric air content to be

A = RT X 2.5 X 79 X 41 dpm/106 ft3, (2)365

where a period of T days elapses between the rains washing out the air mass con-cerned. Taking T to be 6 days on the average, the surface air should contain134 R dpm/106 ft3 in the middle latitudes.The rainfall content will be 4 R dpm/gal for regions with an annual rainfall of

31.5 inches. The concentrations of Sr10 per unit volume for other annual pre-cipitations are to be derived by inverse proportionately, e.g., in Antarctica, wherethe annual precipitation is about one-fourth of 31.5 inches, the Sr90 content of snowshould be given by 16 R,,or

A' = 16 X 1.20et(1/10°+1/40) dpm/gal. (3)In this connection, it is very important to note that regions of frequent rain-fall very probably will receive more Sr90 fallout than will more arid regions.

1961;"ft/'I/d O

FIG. 9.-Predicted fallout rate for Sr'0 formed prior to and duringoperation castle.

Of course, some fallout will be deposited by the surface winds blowing over theleaves of trees and grass. For example, the Naval Research Laboratory10 hasmounted an uncharged platinum screen vertically and held normal to the surfacewinds by a large vane. The deposition is by impact. Two weeks' total collectionswere made, and these gave up to 20 times as much as for gummed papers of thesame area exposed for the same time in the same place. The screen was 80 meshand probably passed about 0.5 X 106 ft.3 in the 2 weeks' exposure. From this itis clear that surface contact with fallout-laden tropospheric air must result in dep-

384 PROC. N. A. S.

Dow

nloa

ded

by g

uest

on

July

1, 2

021

CHEMISTRY: W. F. LIBBY

osition. Further evidence for this is seen in Table 1, where there is seen to beessentially no correlation between the Sr90 contents of soils and the crops of alfalfagrown on them. There obviously must have been very considerable direct deposition on the surface of the plants. Also, the relatively low values for the Sr9"content of surface air found by the Naval Research Laboratory (Table 11), cal-culated with respect to the observed rain content, may very well be due to surfacedeposition by direct contact on tree leaves and grass.The soil content will be the total of all fallout radioactivity, less any natural

weathering processes which serve to remove the fallout Sr90 from the chemically availableform in which plants can assimilate it. Neglecting this latter effect, although, aswe shall see, there is reason to believe that such effects are operative in an importantway, and taking the average for the exchangeable Ca content of soils over theworld (cf. Tables 2 and 3), which is 8 gm/ft2/in. we calculate that the top 2.5 inchesof soil, which in general holds nearly all the Sr90 (Tables 2 and 3), has some 20 gmof exchangeable Ca. Therefore, we predict that in the absence of curative weather-ing effects acting to remove Sr90 from contact with the biosphere, the average Sr90concentration of exchangeable Ca, in MPC units, should be

0.0792 1S =2 X 20 P + R dt] (4)2.2 X 20L J I

where P is the pre-Castle deposition of 0.8 me/mi2 in the middle latitudes and 0.2Mc/mi2 world wide. Thus in the middle latitudes the Sr90 concentration of ex-changeable soil Ca should be given in MPC units by

8 0. 0792 12=2 2 0.8 + 1 2 -/10 dt e-'/40 (5)2.2 X 20

or,

S = [0.0015 + 0.0213(1 - e-t10)]e-'140 (5')This result in terms of Mc/mi2 and MPC units is presented graphically in Figure10. The maximum post-Castle Sr90 soil activity will be expected in about 1970.The present average should be about 0.005 MPC unit for soil with 20 gm. exchange-able Ca/ft2. Those soils of low Ca content can, of course, have a much higherSr90 content for unit amount of exchangeable Ca. Consider, for example, certainareas in Wales near Cardigan (cf. Table 3, Sample No. 54417), where the availableCa/ft2 amounted to only 0.4 gm. and the specific Sr90 activity was found to be0.097 MPC. For this area our analysis would predict a forty fold higher contentthan that given in Figure 10. This, of course, is reflected in higher contents forthe bones of grazing animals.The weathering processes which may operate to fix the fallout Sr90 and make it

unavailable to the biosphere, such as the fixation in massive Ca deposits, areworthy of consideration. Only further investigation will reveal how importantsuch processes are. The present Project Sunshine sampling program includesrepeated sampling of given regions. The data so obtained should disclose anysuch trends. Some data already in hand seem to indicate such effects, but furtherconfirmation is necessary.

VOL. 42, 1956 385

Dow

nloa

ded

by g

uest

on

July

1, 2

021

CHEMISTRY: W. F. LIBBY

: - .0126

6 .000

1900 1960 1980 1976 197. 19,0 1905 1990 190 2000

FIG. 10.-Predicted Sr51' in soil following Castle (Assumed Ca = 20 gms./ft2/2.5")

In addition, palliative mneasures may prove effective. For example, Nervik,Kalkstein, and Libby19 have shown that milk can be purified for radiostrontiumby a treatment which mnay well prove to be quite practical and inexpensive.

B. BIOSPHERE CONTENT OF Sr50

It seems clear that there will be discriminatory barriers of some magnitudeoperating at each of the stages of transfer from the soil into the biosphere. Takingthese stages in general to be

(1) Soil -. Plants,(2) Plants Animals,(3) Animals (milk) -~Humans,

we consider the three corresponding barriers.

1. Soil-to-Plant Transfers.-In Table 12 the Sr90 contents of plants are com-pared with those for the soils on which they grew at a variety of localities justbefore Operation Castle. It seems clear that, on the average, plants have abouttwice the specific Sr90 content relative to Ca that the supporting soils do. It seems

TABLE 12PRE-CASTLE PLANT Sr9° CONTENTS

(Values Are Given in Terms of 1/1,000 MPC Units)Content ofSoil Which

Location Plant Date Reference Plant Content Grew PlantU.S. Alfalfa (Average of Tablel1) 8.9 4.7±+0.4Turkey Alfalfa 10-53 * 2.16±+0.18 1.2±+0.1Cuba Tobacco 4-54 $ 1.7 ± 0.2 (1.6)tNewK Zealand Forage 11-53t 1.17c 0.28 0.21New Zealand Forage 1-54 0.84±+0.10 0.18Chile Forage 11-53 § 0.84p 0.02 0.33*

* E. A. Martell and W. F. Libby, Project Sunshine Bull. No. 10, January 10, 1955.tCalculated from average pre-Castle deposition (Fig. 4a) for latitude of Cuba and assumed average Ca content

of 20 gm/ft2 in top 2.5 inches.E. A. Martell, Project Sunshine Bull. No. 10, 5uppl. 1, March 1, 1955.E. A. Martell, ibid., uppl. 2, June 1, 1955.

very likely that this is due in part to fallout occurring directly on the externalplant surfaces. This is further borne out by the lack of detailed correlation inTable 1 between soil and alfalfa results for eleven middle western United Statesfarms. From this result, we may calculate that about half the totalSr9d contentof alfalfa is due to direct fallout, while for general forage it is probably an even

386 PRoc. N. A. S.

Dow

nloa

ded

by g

uest

on

July

1, 2

021

CHEMISTRY: W. F. LIBBY

higher percentage. This comparison is not quite accurate, because, althoughalmost all the Sr90 is contained in the top 2.5 inches of soil, the Ca, on the otherhand, is available to the entire depth of the root system. It is also recognizedthat the vertical distribution of the Ca may be nonuniform. There is no reasonto expect preferential assimilation of Sr from the soil relative to Ca; hence theonly other explanation for the data in Table 12 is direct fallout. As remarked inSection I in the paragraphs on rain and air-filter data, there are two mechanismsfor direct fallout of the ultra-finely divided particulate matter carrying the Sr90in the stratospheric reservoir-rainout and contact deposition after the fallouthas entered the troposphere. Rainout appears to be the principal mechanism,though it has been demonstrated by the Naval Research Laboratory,20 as men-tioned earlier, that an 80-mesh screen mounted vertically to prevailing surfacewinds can gather more fallout than falls out directly on the average by all mech-anisms. It is not clear, however, to what degree foliage acts in this way, butthe probability seems high that the effect is relatively minor and that rainout,followed by drying of raindrops, is the main way in which the foliage surfacesgather fallout.On this basis we conclude that total fallout in arid regions should be appreciably

lower than in areas with normal rainfall. This effect seems to be borne out by thedata available now, though more definitive experiments are needed. It probablyfollows also that regions subject to seasonal rainfall rather than relatively uniformprecipitation all year should show less fallout for the same total annual rainfall.Also, regions subject to frequent morning fogs may be particularly high in totalfallout. The Sr90 probably will enter the troposphere at relatively uniform rates,but the chance of precipitation will depend strongly on the local weather.Of course, rainfall is necessary to plant growths, so that plants are certain to

gather some fallout. However, for regions of low rainfall where irrigation isused-such as the Imperial Valley in California-the fallout content of the cropsshould be particularly low, for, as shown in Table 7, rivers are nearly free of fallout,since the soil purifies the runoff water before it reaches the rivers. Similarly,reservoirs and lakes will be low relative to rain because of dilution and the im-portance of runoff water from surrounding watersheds in replacing evaporative andwithdrawal losses. It is also well to note that the ordinary water purificationprocesses are effective in removing an appreciable fraction of the radiostron-tium.The by-passing of soil entirely, which occurs in the direct fallout on plant sur-

faces, of course means that the retarding effects of high Ca contents in soil areinoperative, and cattle grazing on such foliage may show little correlation in theSr90 contents with the soil Sr90 activity for this reason. This appears, from thedata in Table 1, to be true in the United States Midwest.

It is clear, however, that washing may reduce the level of fallout externallycarried by plants, though direct leaf absorption will be expected to occur ratherrapidly for water-soluble fallout. For the megaton weapons fired in the Pacific,the bulk of the fallout resides on particles of CaO or Ca(OH)2 or mixtures of CaO,Ca(OH)2, and CaCO3 made by the great heat of the fireball acting on the coral ofthe islands and sea floor in the firing areas.2' A large amount also is carried onNaCl particles. This material, therefore, should be quite water-soluble and should

VOL. 42, 1956 387

Dow

nloa

ded

by g

uest

on

July

1, 2

021

CHEMISTRY: W. F. LIBBY

be rapidly absorbed into the leaves. Washing therefore probably will not beparticularly effective for the world-wide fallout, which derives from the Pacifictests. From weapons fired in the air, the particles probably will consist of lesssoluble oxides and therefore are more likely to wash off of plant surfaces beforebeing absorbed.

Menzel,22 of the United States Department of Agriculture, grew cowpeas on42 American soils to which equal amounts of bomb debris had been added. Avail-able Ca ranged from 0.7 to 48 milliequivalents of Ca/100 gm of soil. The Sr90/Caratio of the plants was approximately inversely proportional to the available Cain the soil over the full range of Ca availability. In another set of experiments ona particular type of soil (Evesboro) to which known amounts of Sr90 had been addedat two carrier levels, the results listed in Table 13 were found. The distribution

TABLE 13PLANT ASSIMILATION OF Sr9° FROM SOIL

(Sr/Ca) SoilCrop (By Equivalents) ksr klC

Barley~ ~ 0.017 0.45 0.020Barley 10.0017 0.39 0.022Buckwheat 10017 0.49 0.023~0.0017 0.43 0.028

f0.017 0.53 0.057Cowpeas 10.0017 0.37 0.053

factor, k7r, defined as (Sr/Ca)plant/(Sr/Ca)soil, indicates the discriminationwhich the plant makes between Sr and Ca uptake. Similar tests were made for Ba.By combining these data, Menzel concluded, as shown in Table 13, that the

average Sr uptake from American soils was best fitted by a distribution factor ofk~r = 0.36. This average probably will apply world-wide about as well.

2. Plant-to-Animal Transfer.-The Sr90 contained in grass and foliage eatenby grazing animals will be retained to an extent dependent on the metabolismof the animal. For example. for a 1-year-old steer23 30 per cent of the Sr90 fedorally was retained with essentially no discrimination relative to Ca. There ap-pears to be a higher retention, approaching 90 per cent, for intravenous injectionof young rats.23 High Ca diets reduce the Sr90 uptake for rats, adult rats take upabout 16 per cent of the ingested Sr90 on the same low Ca diet for which the youngrats took up 73 per cent.23Comar24 has performed experiments on cows in which the Sr/Ca ratio in feed,

blood, and milk was measured under equilibrium conditions. Typical relativevalues were 1.0, 0.37, and 0.13, respectively, thus indicating a relative lowering ofthe Sr90 content of the milk relative to the feed. This is borne out by direct ob-servation on the fallout, as seen in Table 1, where for the Chicago milkshed themilk averaged 0.0014 MPC, while the alfalfa fed averaged 0.0089 MPC.

3. Milk-to-Human Transfer.-This ability of the cow to reduce the Sr90 in themilk relative to the feed is important as a barrier to human ingestion of this falloutradioactivity. With it in mind, we can expect that human Sr90 burdens should be20 per cent or less (for older people) of the plant contents and about equal to themilk and cheese levels if the entire Ca in the body were assimilated at a given Sr90content of the milk.

388 PROC. N. A. S.

Dow

nloa

ded

by g

uest

on

July

1, 2

021

CHEMISTRY: W. F. LIBBY

Since the Sr90 content of the whole biosphere is continually rising, about as shownin Figure 10, the average Sr90 content of milk should rise in a similar manner.Therefore, the intake of Sr90 by humans is steadily increasing, as shown in Figuresla, lb, Ic, and Id. The data in Table 3 and Figures Ic and id show that the milklevel at the time of Operation Castle averaged about 0.001 MPC, peaking in themiddle latitudes, as did the soil assays (cf. Fig. 4a). This value shows a ratio ofabout 0.7 for the milk level to the average soil level for the top 2.5 inches of soil with20 gm. of contained available Ca/ft2. This high value probably is due to leafpickup of fallout. Taking this as a general result, we predict the world valuesfor milk and cheese at about 70 per cent of the soil values given in Figure 10, plusainv local fallout. This would mean that we would expect average foreign milkand cheese samples to show about 0.004 MPC at the present time.The human bone, of course, is formed during the growing process, so that the

Sr90 content of children should be higher than for adults. The data2 8-12, 25 verifythis prediction; however, for newborn babies there is less Sr90 than corresponds tothe milk, showing the retarding effect of the mother's older Ca pool. Childrenseem to carry Sr90 approximately equal to the average level of the milk duringthe period of their lives. Adults decrease in Sr90 concentration with age as ex-pected. It seems that the adults of the future will have Sr90 levels correspondingto the milk levels during their lifetime weighted according to their rate of growth.For example, the years from 10 to 18 will be most important for men and from6 to 12 for women. Foreign children born now, according to Figure 10, shoulddevelop about 0.011 MPC unit during their lives. Children born now in theUnited States will develop a somewhat higher level, due to somewhat higher milklevels.

I "World-wide Effects of Atomic Weapons: Project Sunshine," Rand Corporation, R-251-AEC,August 6, 1953.

2 W. R. Eckelmann and J. L. Kulp, Project Sunshine Bulletin, October 15, 1955.' S. Kaufmann and W. F. Libby, "The Natural Distribution of Tritium," Phys. Rev., 93,

1337, 1954.4 H. von Buttlar and W. F. Libby, "Natural Distribution of Cosmic Ray Produced Tritium.

II," J. Inorg. and Nuclear Chem., 1, 75, 1955.6L. A. Currie, W. F. Libby, and R. Wolfgang, "Tritium Production by High Energy Protons,"

Phys. Rev., 101, 1557, 1956.6 F. Begemann and W. F. Libby, "Tritium Content of Surface Water, March 1954 to January

1956" (in press).I Personal communication from Dr. Lester Machta, United States Weather Bureau.8 E. A. Martell and W. F. Libby, Project Sunshine Bulletin No. 10, January 10, 1955.9 E. A. Martell, Project Sunshine Bulletin No. 10, Suppl. 1, March 1, 1955.

10 E. A. Martell, ibid., Suppl. 2, June 1, 1955.1' E. A. Martell, ibid., Suppl. 3, September 1, 1955.12 E. A. Martell, Project Sunshine Bulletin No. 11, December 1, 1955.13 G. H. Wilkening, "Natural Radioactivity as a Tracer in the Sorting of Aerosols According to

Mobility," Rev. Sci. Instr., 23, 13, 1952.14 I. H. Blifford, L. B. Lockhart, Jr., and H. B. Rosenstock, "On the Natural Radioactivity in

the Air," J. Geophys. Research, 57, 499, 1952.15 0. Haxel and G. Schumann, "Selbstreinigung der Atmosphare," Z. Physik, 142, 127-132,

1955.18 P. E. Damon and P. K. Kuroda, "On the Natural Radioactivity of Rainfall," Trans. Am.

Geophys. Union, 35, 208, 1954.17 "World-wide Fallout from Operation Castle," NYOO Report NYO-4621, January 21, 1955.

V!OL. 42, 1956 389)

Dow

nloa

ded

by g

uest

on

July

1, 2

021

CHEMISTRY: W. F. LIBBY

18 Project Sunshine: Progress from September 1953 to January 4, 1954, NYOO ReportNYO-4571, January 8, 1954.

19 W. E. Nervik, M. I. Kalkstein, and W. F. Libby, "Purification of Milk for Calcium andStrontium with Dowex X-50W Resin,," University of California Radiation Laboratory Report!No. UCRL-2674, August 12, 1954.

20 I. H. Blifford, Jr., and H. B. Rosenstock, "Fallout Dosages at Washington, D.C.," NavalResearch Laboratory Report NRI4654, November 3, 1955; I. H. Blifford, Jr., L. B. LockhartJr., and R. A. Baus, "Relationship between the Air Concentration of Radioactive Fission Prod-ucts and Fallout," Naval Research Laboratory Report NRL-4607, November 4, 1955; R. A.Baus, I. H. Blifford, Jr., and L. B. Lockhart, Jr., "Radioactivity of the Air," Naval ResearchLaboratory Report NRL-4509, March, 1955.

21 "Fallout Symposium," AFSWP-895, January, 1955.22 R. G. Menzel and I. C. Brown, "Leaching of Fallout and Plant Uptake of Fallout," bimonthly

report from United States Department of Agriculture, March-April, 1953.23 United States Atomic Energy Commission, Division of Biology and Medicine Report, No.

100A July, 1954.24 C. Comar, "Agricultural Research Program-Semiannual Progress Report for July 1-

December 31, 1953," Report ORO-110, April, 1954, and C. L. Comar and R. H. Wasserman,"Radioisotopes in the Study of Mineral Metabolism," in Progress in Nuclear Energy, Vol. 5,("Biological Sciences and Medicine" [London: Pergamon Press, Ltd., 1956]) (in press).

25 H. L. Volchok and J. L. Kulp, Project Sunshine Annual Progress Report, March 15, 1955.26 Sunshine Report for May and June, NYOO Report NYO-4653, July 5, 1955.27 Sunshine Report for July and August, NYOO Report NYO-4661, September 16, 1955.28 Sunshine Report, August, 1954, NYOO, HASL-S-2 (NYO4656), November 12, 1954.29 NYOO Report NYO-4623, January 18, 1955.

PROC. N. A. S.-39}0

Dow

nloa

ded

by g

uest

on

July

1, 2

021